MuPeG—The Multiple Person Gait Framework
Gait recognition is being employed as an effective approach to identify people without requiring subject collaboration. Nowadays, developed techniques for this task are obtaining high performance on current datasets (usually more than <inline-formula> <math display="inline"> &l...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2020-03-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/20/5/1358 |
| _version_ | 1828147000211668992 |
|---|---|
| author | Rubén Delgado-Escaño Francisco M. Castro Julián R. Cózar Manuel J. Marín-Jiménez Nicolás Guil |
| author_facet | Rubén Delgado-Escaño Francisco M. Castro Julián R. Cózar Manuel J. Marín-Jiménez Nicolás Guil |
| author_sort | Rubén Delgado-Escaño |
| collection | DOAJ |
| description | Gait recognition is being employed as an effective approach to identify people without requiring subject collaboration. Nowadays, developed techniques for this task are obtaining high performance on current datasets (usually more than <inline-formula> <math display="inline"> <semantics> <mrow> <mn>90</mn> <mo>%</mo> </mrow> </semantics> </math> </inline-formula> of accuracy). However, those datasets are simple as they only contain one subject in the scene at the same time. This fact limits the extrapolation of the results to real world conditions where, usually, multiple subjects are simultaneously present at the scene, generating different types of occlusions and requiring better tracking methods and models trained to deal with those situations. Thus, with the aim of evaluating more realistic and challenging situations appearing in scenarios with multiple subjects, we release a new framework (MuPeG) that generates augmented datasets with multiple subjects using existing datasets as input. By this way, it is not necessary to record and label new videos, since it is automatically done by our framework. In addition, based on the use of datasets generated by our framework, we propose an experimental methodology that describes how to use datasets with multiple subjects and the recommended experiments that are necessary to perform. Moreover, we release the first experimental results using datasets with multiple subjects. In our case, we use an augmented version of TUM-GAID and CASIA-B datasets obtained with our framework. In these augmented datasets the obtained accuracies are <inline-formula> <math display="inline"> <semantics> <mrow> <mn>54.8</mn> <mo>%</mo> </mrow> </semantics> </math> </inline-formula> and <inline-formula> <math display="inline"> <semantics> <mrow> <mn>42.3</mn> <mo>%</mo> </mrow> </semantics> </math> </inline-formula> whereas in the original datasets (single subject), the same model achieved <inline-formula> <math display="inline"> <semantics> <mrow> <mn>99.7</mn> <mo>%</mo> </mrow> </semantics> </math> </inline-formula> and <inline-formula> <math display="inline"> <semantics> <mrow> <mn>98.0</mn> <mo>%</mo> </mrow> </semantics> </math> </inline-formula> for TUM-GAID and CASIA-B, respectively. The performance drop shows clearly that the difficulty of datasets with multiple subjects in the scene is much higher than the ones reported in the literature for a single subject. Thus, our proposed framework is able to generate useful datasets with multiple subjects which are more similar to real life situations. |
| first_indexed | 2024-04-11T20:54:10Z |
| format | Article |
| id | doaj.art-db0ad545b6f544fca40f0e6f5210fc91 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-04-11T20:54:10Z |
| publishDate | 2020-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-db0ad545b6f544fca40f0e6f5210fc912022-12-22T04:03:45ZengMDPI AGSensors1424-82202020-03-01205135810.3390/s20051358s20051358MuPeG—The Multiple Person Gait FrameworkRubén Delgado-Escaño0Francisco M. Castro1Julián R. Cózar2Manuel J. Marín-Jiménez3Nicolás Guil4Department of Computer Architecture, University of Málaga, 29071 Málaga, SpainDepartment of Computer Architecture, University of Málaga, 29071 Málaga, SpainDepartment of Computer Architecture, University of Málaga, 29071 Málaga, SpainDepartment of Computer Science and Numerical Analysis, University of Córdoba, 14071 Córdoba, SpainDepartment of Computer Architecture, University of Málaga, 29071 Málaga, SpainGait recognition is being employed as an effective approach to identify people without requiring subject collaboration. Nowadays, developed techniques for this task are obtaining high performance on current datasets (usually more than <inline-formula> <math display="inline"> <semantics> <mrow> <mn>90</mn> <mo>%</mo> </mrow> </semantics> </math> </inline-formula> of accuracy). However, those datasets are simple as they only contain one subject in the scene at the same time. This fact limits the extrapolation of the results to real world conditions where, usually, multiple subjects are simultaneously present at the scene, generating different types of occlusions and requiring better tracking methods and models trained to deal with those situations. Thus, with the aim of evaluating more realistic and challenging situations appearing in scenarios with multiple subjects, we release a new framework (MuPeG) that generates augmented datasets with multiple subjects using existing datasets as input. By this way, it is not necessary to record and label new videos, since it is automatically done by our framework. In addition, based on the use of datasets generated by our framework, we propose an experimental methodology that describes how to use datasets with multiple subjects and the recommended experiments that are necessary to perform. Moreover, we release the first experimental results using datasets with multiple subjects. In our case, we use an augmented version of TUM-GAID and CASIA-B datasets obtained with our framework. In these augmented datasets the obtained accuracies are <inline-formula> <math display="inline"> <semantics> <mrow> <mn>54.8</mn> <mo>%</mo> </mrow> </semantics> </math> </inline-formula> and <inline-formula> <math display="inline"> <semantics> <mrow> <mn>42.3</mn> <mo>%</mo> </mrow> </semantics> </math> </inline-formula> whereas in the original datasets (single subject), the same model achieved <inline-formula> <math display="inline"> <semantics> <mrow> <mn>99.7</mn> <mo>%</mo> </mrow> </semantics> </math> </inline-formula> and <inline-formula> <math display="inline"> <semantics> <mrow> <mn>98.0</mn> <mo>%</mo> </mrow> </semantics> </math> </inline-formula> for TUM-GAID and CASIA-B, respectively. The performance drop shows clearly that the difficulty of datasets with multiple subjects in the scene is much higher than the ones reported in the literature for a single subject. Thus, our proposed framework is able to generate useful datasets with multiple subjects which are more similar to real life situations.https://www.mdpi.com/1424-8220/20/5/1358gait recognitiongait frameworkgait datasetmultiple subjectsaugmented dataset |
| spellingShingle | Rubén Delgado-Escaño Francisco M. Castro Julián R. Cózar Manuel J. Marín-Jiménez Nicolás Guil MuPeG—The Multiple Person Gait Framework Sensors gait recognition gait framework gait dataset multiple subjects augmented dataset |
| title | MuPeG—The Multiple Person Gait Framework |
| title_full | MuPeG—The Multiple Person Gait Framework |
| title_fullStr | MuPeG—The Multiple Person Gait Framework |
| title_full_unstemmed | MuPeG—The Multiple Person Gait Framework |
| title_short | MuPeG—The Multiple Person Gait Framework |
| title_sort | mupeg the multiple person gait framework |
| topic | gait recognition gait framework gait dataset multiple subjects augmented dataset |
| url | https://www.mdpi.com/1424-8220/20/5/1358 |
| work_keys_str_mv | AT rubendelgadoescano mupegthemultiplepersongaitframework AT franciscomcastro mupegthemultiplepersongaitframework AT julianrcozar mupegthemultiplepersongaitframework AT manueljmarinjimenez mupegthemultiplepersongaitframework AT nicolasguil mupegthemultiplepersongaitframework |